OMAHA, Neb. (KMTV) — Scams are nothing new, but the rapid growth of artificial intelligence is making scams even more difficult to spot.
3 News Now's Zach Williamson visited University of Nebraska Omaha professor and computer science expert — Dr. Victor Winter — to help you keep yourself and your money safe.
- Dr. Winter teaches two generative AI courses at the university and works in the ever-changing landscape of AI every day.
- Reports of generated AI enabled scams surged by 456 percent between May of 2024 and April of 2025 compared to the previous year.
- The biggest piece of advise: never click on a link without confirmation from the person or organization who sent it. The same goes for a phone call — hang up and confirm.
BROADCAST TRANSCRIPT:
It's predicted that generative AI-driven fraud losses in the US alone could exceed $40 billion by 2027.
“In the past, if you wanted to do a spear-phishing attack or send someone an email you had sort of two choices. You can do a generic one, like the Nigerian prince scam, or you could do your homework and get a dossier on this person and craft something, and skills are required to do this right,” Dr. Winter explained.
“Now, it’s easy for anyone to do. So targeted attacks can be done at scale and they’re better than the best humans.”
He says his general rule is to never click on a link without confirmation from the person or organization who sent it. The same goes for a phone call — hang up and confirm. Confirming who the person really is that is talking to you can be very tricky.
“The way we understand the world is through our eyes and ears, right? Now this has called that into question,” Dr. Winter said.
He says duplicating someone’s voice now takes just ten seconds, and it is possible to use picture animation to match the voice to a person’s face.
“This is a webpage I created that lets you explore how easy it is for an AI to create a targeted spear-phishing attack on an individual.”
It's for educational purposes only. He created fake people with fake Facebook and Linked-In accounts. Showing how AI can pull any information on you that’s on the internet and social media, as well as build a persona.
In this example, you can come across as a family member, a friend, a co-worker, or a boss. However, he showed us the AI can build personas to mirror really anything, like generations (boomer, millennial, Gen Z).
Then, it writes a message to fool you.
“You have to introduce in this equation information that the AI cannot know. So, this is like a safe word. Like if you say, 'what city was I born in?’ That may show up somewhere. It has to be something the AI cannot not know.”
Dr. Winter says — like AI itself — scams will only become more sophisticated. I’ve included a video below with a bit of our conversation on the rapid growth of AI.
